临床大数据挖掘与人工智能建模云平台的研究与构建

Research and construction of the cloud platform for clinical big data mining and artificial intelligence modeling

目的:为解决临床研究中分析、建模及项目管理的相关问题,整合多种统计及机器学习方法,构建智能临床研究分析及AI建模云平台。方法:采用Django作为Web应用框架,Nginx负载均衡,使用R语言实现生存分析、COX回归分析等传统临床统计方法,同时集成了XGBoost、SVM、随机森林等多种机器学习算法,构建了一键自动人工智能建模的云平台。结果:平台目前已完成搭建,集数据治理、医学统计学、机器学习建模、项目管理、流程追溯于一体,可以一站式全流程完成临床研究分析建模。结论:本平台有助于减少临床医生开展临床研究时数据整理、分析及建模的时间成本和经济成本,同时提供了技术和方法支撑,有效提升了临床研究的效率和质量。

Objective To solve the problems related to analysis,modeling and project management in clinical research,integrate several statistical and machine learning methods,and build the intelligent clinical research analysis and AI modeling cloud platform.Methods Django was used as the Web application framework,Nginx has a balanced load,and R language was used to realize the traditional clinical statistical methods such as survival analysis and COX regression analysis,while a variety of machine learning algorithms such as XGBoost,SVM and random forest were integrated to build a cloud platform for one-click automatic artificial intelligence modeling.Results At present,the platform has been built.It integrates data governance,medical statistics,machine learning modeling,project management and process tracing,and can complete the clinical research,analysis and modeling in the whole process.Conclusion This platform helps to reduce the time cost and economic cost of data sorting,analysis and modeling when clinicians do the clinical research,and provides technical and methodological support and effectively improves the efficiency and quality of clinical research.